Novel Bioactive Hybrid Compound Dual Targeting Estrogen Receptor and Histone Deacetylase for the Treatment of Breast Cancer

Article abstract of DOI:10.1021/acs.jmedchem.5b00099

A strategy to develop chemotherapeutic agents by combining several active groups into a single molecule as a conjugate that can modulate multiple cellular pathways may produce compounds having higher efficacy compared to that of single-target drugs. In th

Full text of DOI:10.1021/acs.jmedchem.5b00099

Subscriber access provided by NEW YORK UNIV
Article
Novel Bioactive Hybrid Compounds Dual Targeting Estrogen
Receptor and Histone Deacetylase for Treatment of Breast Cancer
Chu Tang, Changhao Li, Silong Zhang, Zhiye Hu, Jun Wu, Chune Dong, Jian Huang, and Hai-Bing Zhou
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 20 May 2015
Downloaded from http://pubs.acs.org on May 21, 2015
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
Novel Bioactive Hybrid Compounds Dual Targeting Estrogen Receptor and
Histone Deacetylase for Treatment of Breast Cancer
7
8
9
Chu Tang,a‡ Changhao Li,b‡ Silong Zhang,a Zhiye Hu,a Jun Wu,b Chune Dong,a* Jian
Huang,b* and HaiꢀBing Zhoua*
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a State Key Laboratory of Virology;
Key Laboratory of Combinatorial Biosynthesis and Drug Discovery
(Wuhan University), Ministry of Education;
Wuhan University School of Pharmaceutical Sciences, Wuhan, 430071,
China;
b College of Life Sciences, Wuhan University, Wuhan, 430072, China
‡ These two authors contributed equally to this work
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 58
1
2
3
4
ABSTRACT
5
6
7
8
A strategy to develop chemotherapeutic agents by combining several active groups
into a single molecule as a conjugate that can modulate multiple cellular pathways
may produce compounds having higher efficacy compared to singleꢀtarget drugs. In
this paper, we describe the synthesis and evaluation of an array of dualꢀacting ER and
histone deacetylase inhibitors. These novel hybrid compounds combine an indirect
antagonism structure motif of ER (OBHS, oxabicycloheptene sulfonate) with the
HDAC inhibitor suberoylanilide hydroxamic acid (SAHA). These OBHSꢀHDACi
conjugates exhibited good ER binding affinity and excellent ERα antagonistic activity,
and they also exhibited potent inhibitory activities against HDACs. Compared with
the approved drug tamoxifen, these conjugates exhibited higher antitumor potency in
ERαꢀpositive breast cancer cells (MCFꢀ7). Moreover, these conjugates not only
showed selective anticancer activity that was more potent against MCFꢀ7 cells than
DU 145 (prostate cancer), but they had no toxicity towards normal cells.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 58
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
Breast cancer is the most common cancer among women, and approximately
230,000 breast cancer cases are reported per year in the United States, of which
approximately 13,000 are diagnosed in women aged less than 40 years.1 The estrogen
receptor (ER), which is a ligandꢀregulated transcription factor that regulates many
physiological and pathological processes, plays a predominant role in breast cancer
growth.2, 3 Therefore, ER is regarded as important pharmaceutical target for the
treatment of breast cancer, and development of ER ligands has emerged as active
study field in the fight against breast cancer. Many of these ligands have been
developed into hormone agents, which often have mixed agonistꢀantagonist and
tissueꢀselective activities. Some of these agents have been termed selective estrogen
receptor modulators (SERMs),4, 5 which are the most recently approved class of first
line drugs for the treatment of breast cancer. Tamoxifen is the first clinically used
SERM for the prevention and treatment of breast cancer.6, 7
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Although current SERMs have clear advantages in the treatment of
hormoneꢀresponsive breast cancer, they retain some disadvantages. For instance,
tamoxifen therapy is ineffective for ER(ꢀ) and tripleꢀnegative breast (TNB) cancers,
and roughly half of ER(+) tumors are insensitive or gain resistance, losing their
response to continued tamoxifen therapy.8ꢀ10 Thus, much effort has been undertaken
to develop an “ideal SERM”, one that is more effective or one towards which
resistance does not develop.11 One strategy to develop such SERMs involves
combining two bioactive drugs into a single molecule, forming a conjugate that can
interact with two relevant components of breast cancer process and thus might possess
enhanced therapeutic activity.12
In taking this approach, the attachment of antiꢀtumor agents to the functional
groups of the ER ligands, estradiol or tamoxifen, has been used to form conjugates
that exhibited enhanced pharmacological properties. For example, conjugates have
been made with an intercalating agent (E2ꢀellipticine conjugate 1, Figure 1),13 an
alkylating agent (E2ꢀchlorambucil conjugate 2, Figure 1),14 or an antimitotic
(tamoxifenꢀdoxorubicin conjugate 3, Figure 1).15 Although these conjugates have very
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 58
1
2
3
4
5
6
7
8
potent antiꢀcancer activity, because of their extremely low ER binding affinity, they
lack the cellꢀtype or targetꢀtissue selectivity, which severely limits their clinical utility.
So far, the estrogen receptor targeting agents used to form these types of conjugates
9
have mostly been estradiol or tamoxifen; other ER ligands have rarely been chosen.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
N
OH
O
H
N
N
H
O
N
H
1
HO
OH
Cl
N
Cl
H
O
N
HO
O
N
5
H
O
OH
O
2
HO
O
OH
O
O
OH
O
N
O
H
O
N
O
H3C
HO
HN
HO
HN
3
HO
Figure 1. (1) E2ꢀellipticine conjugate; (2) E2ꢀchlorambucil conjugate; (3)
tamoxifenꢀdoxorubicin conjugate.
As part of our ongoing interest in the development of ER ligands having
therapeutic efficacy on breast cancer,16ꢀ18 we have focused on the preparation and
evaluation of novel ER ligands having a more threeꢀdimensional character. Among
them, exoꢀ5,6ꢀbisꢀ(4ꢀhydroxyphenyl)ꢀ7ꢀoxabicyclo[2.2.1]heptꢀ5ꢀeneꢀ2ꢀsulfonic acid
phenyl ester (OBHS), was one of the best compounds, exhibiting modest ER subtype
selectivity, with relative binding affinity (RBA) values of 9.3% and 1.7% (data from
reference 16) for ERα and ERβ, respectively (RBA[estradiol] = 100%), and was a
partial antagonist on both ERs (Figure 2).16 Although OBHS bears no structural
resemblance to other estrogen antagonists, analysis of the Xꢀray crystal structure of
the complex of ERαꢀLBD with OBHS indicates that its partial antagonist character
was achieved by indirectly modulating the critical switch helix in the ER ligand
binding domain, helix 12, by interactions with helix 11. The crystal structure also
4
ACS Paragon Plus Environment

Page 5 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
shows that the phenyl sulfonate group of OBHS plays an important role in producing
the partial antagonist activity of OBHS: The sulfonate phenol binds between helices 8
and 11, and by making a strong steric clash with helix 11, it displaces His524 outward
and repositions helix 11 so that it indirectly modulates the orientation of the helix 12,
eventually destroying the surfaceꢀbound coactivator binding site. Furthermore, in
contrast to the mechanism action of tamoxifen that directly relocates helix 12 through
its interaction with the bulky and basic side chain, OBHS has two 4ꢀhydroxyphenyl
substituents, one which of which mimics the hydroxyphenyl of estradiol, engaging in
strong hydrogen bond with E353; the other one points in the E2 11β direction, but it is
not long enough to interact directly with helix 12. Therefore, this second phenol group
can be replaced or modified with a range of functional group to enhance the
antagonist properties of these oxabicyclo[2.2.1]heptene core ligands.19
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
With regard to the design of our conjugates, it is notable that estrogen signaling
requires the displacement of certain proteins, for example, HDACs, from corepressor
complexes, and the recruitment of coactivator proteins to transcription complexes
containing liganded ER.20 Solid evidence supports the direct interaction of HDACs
with corepressor proteins and ER in silenced nuclear transcription complexes in the
cell nucleus, for example, the interaction of HDAC1 with the activation function 2
(AF2) and DNA binding domain (DBD) of ERα.21 In ER(+) cells, knockdown of
HDAC1 and HDAC inhibition by either trichostatin A (TSA), valproic acid (VPA), or
butyric acid, can decrease ERα levels.22ꢀ25 Thus, because the combination of SERM
activity with HDAC inhibitor therapy requires both drug components to be in close
proximity in the cell nucleus, the concept of an HDAC inhibitor combined with a
SERM in a single molecule hybrid or conjugate is thus proposed.26 A recent phase II
clinical study also revealed that the use of a combination of the HDAC inhibitor,
SAHA, and the SERM, tamoxifen, as separate agents, could increase the efficacy of
tamoxifen.27
Thus, to address the need for more effective drugs for breast cancer, we chose the
clinically effective HDACi (SAHA) as the second component of our conjugate,
hypothesizing that the incorporation of SAHA into unique ER ligands might produce
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 58
1
2
3
4
5
6
7
8
9
new bifunctional hybrid agents having improved efficacy and selectivity while
retaining high affinity for ER. In light of our interest in ER ligands, we expanded the
diversity of OBHS by equipping this ligand in two ways with the histone deacetylase
inhibitor (HDACi) SAHA so as to confer selective antiꢀtumor activity against breast
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
cancers (Figure 2).
HO
O
O
S O
H
O
N
1
NHOH
6
O
O
5
3
SAHA (Vorinostant)
HO
OBHS
ER = 9.3% (RBA, Estradiol = 100%)
ER = 1.7%
= 5.5
Design of OBHS-HDACi conjugates
(HONH) OH
H
N
HO
O
O
O
O
S O
O
6
O
O
S O
O
O
O
N
OMe (OH, NHOH)
R
6
H
Series I
Series II
HO
HO
Figure 2. Design of dualꢀacting OBHSꢀHDACi conjugates.
Herein, we describe the design and biological evaluation of novel OBHSꢀHDACi
conjugates, which can be prepared conveniently by a DielsꢀAlder reaction of a furan
with an appropriate dienophile. The OBHSꢀSAHA conjugates of this design not only
have threeꢀdimensional topology, but also have second pharmacologically HDACi
unit. Thus they expand our exploration of novel estrogen receptor ligand. For
comparison, the ester and carboxylic acid analogues (by replacing the hydroxamic
acid (ꢀCONHOH) group with (ꢀCOOH) or methyl ester (ꢀCOOMe)) were also
prepared and evaluated for their ER binding affinity, estrogen responsive
elementꢀdriven transcriptional activity, HDAC inhibition activity, cell
antiproliferative activity and selectivity etc.
RESULTS AND DISCUSSION
Chemical Synthesis. All designed OBHSꢀHDACi conjugates were synthesized by a
DielsꢀAlder reaction of 3,4ꢀbis(4ꢀhydroxyphenyl)furan 4 (cf. Scheme 3A) or furan
derivative 12 with various dienophiles. The 3,4ꢀbis(4ꢀhydroxyphenyl)furan 4 was
6
ACS Paragon Plus Environment

Page 7 of 58
Journal of Medicinal Chemistry
1
2
methodology.16
3
4
prepared
according
to
our
previously
developed
5
6
7
8
8ꢀ(4ꢀ(4ꢀ(4ꢀHydroxyphenyl)furanꢀ3ꢀyl)ꢀphenylamino)ꢀ8ꢀoxooctanoic acid 12 was the
key intermediate in the synthesis of OBHSꢀHDACi conjugates, and it was synthesized
by the general route depicted in Scheme 1. Treatment of 4ꢀmethoxyacetophenone 5
with Nꢀbromosuccinimide in the presence of pꢀtoluenesulfonic acid provided
αꢀbromoꢀ4ꢀmethoxyacetophenone 6 in 95% yield. 8 was formed by reaction of 6 with
4ꢀaminophenylacetic acid 7 in the presence of triethylamine in acetonitrile, followed
by an Aldol reaction to give 9, which was demethylated with BBr3 to afford
butenolide 10. Diisobutylaluminum hydride reduction of 10 at ꢀ78 °C gave, after
acidic workup with 4% H2SO4, the furan 11, which was allowed to react with suberic
anhydride to obtain the key intermediate 12.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The synthesis of various vinyl sulfonates 16aꢀb, 19aꢀr was accomplished by the
reaction of 2ꢀchloroethanesulfonyl chloride with substituted phenols under basic
reaction conditions, as shown in Scheme 2. Vinyl sulfonates 16aꢀb could be
synthesized by a threeꢀstep procedure from commercially available suberic acid via
suberic anhydride. First, suberic anhydride 14 was prepared from a mixture of suberic
acid 13 with acetic anhydride.28 Then, suberic anhydride 14 was condensed with the
corresponding aminophenol to generate compounds 15aꢀb, which were subsequently
reacted with 2ꢀchloroethanesulfonyl chloride to yield the vinyl sulfonates 16aꢀb
(Scheme 2A). The various commercially available substituted phenols were reacted
with 2ꢀchloroethanesulfonyl chloride to yield the corresponding vinyl sulfonates
19aꢀr (Scheme 2B). The synthesis of OBHSꢀHDACi conjugates was effectively
accomplished by a DielsꢀAlder reaction of furan 4 or 12 with various vinyl sulfonates
(Scheme 3); the results were summarized in Table 1.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 58
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 1. Synthesis of furan derivative 12. Reagents and conditions: (a) NBS (1.2
equiv.), pꢀTsOH (0.2 equiv.), CHCl3, rt, 12h; (b) Et3N, CH3CN, rt, 12h; (c) NaH,
DMSO, rt, 3h; (d) BBr3, CH2Cl2, ꢀ20 °C, 12h; (e) DIBALꢀH, THF, ꢀ78 °C, 8h; (f)
suberic anhydride , THF, rt, 2h.
Scheme 2. Synthesis of dienophiles 16aꢀb, 19aꢀr. Reagents and conditions: (a) acetic
anhydride, 150 °C, 1h; (b) 4ꢀaminophenol or 3ꢀaminophenol, THF, rt, 1h; (c) (i)
2ꢀchloroethanesulfonyl chloride, Et3N, THF, 0 °C, 12h; (ii) MeOH, rt, 12h (d) Et3N,
CH2Cl2, 0 °C, 12h; (e) BBr3, CH2Cl2, ꢀ20 °C, 12h.
The DielsꢀAlder reaction of vinyl sulfonates with furans (4 or 12) went very
smoothly, and the yields of the products were generally good. Also, it is noteworthy
that high stereoselectivity was observed in the reaction of furans with dienophiles; as
we described previously,16 the exo products predominated in the DielsꢀAlder reaction.
The compound 4 was a symmetrical furan; therefore, the DielsꢀAlder products 20aꢀb,
21aꢀb and 22 were studied single isomers, and among them, 21a showed the highest
binding affinity (see below). The DielsꢀAlder reactions of vinyl sulfonates with the
8
ACS Paragon Plus Environment

Page 9 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
unsymmetrical furan 12 produces the conjugates 23aꢀr, 24a-i, which were studied as a
mixture of regioisomers that, despite our best efforts, could not be separated. The low
affinity of these conjugates compared to OBHS make this less of an issue. As
mentioned above and reported previously, the isolated products are the exo
diastereomers; they are also racemates.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
During our work on OBHSꢀcore ER ligands, we found that the disposition and size
of the substituents on phenyl ring of the OBHS sulfonate moiety were important
determinants of the binding affinity and selectivity of these compounds. First, we
introduced the SAHA group on the phenyl sulfonate unit (Scheme 3A, Series I).
Reaction of 4 with 16aꢀb afforded the ester 20aꢀb, and hydrolysis in basic medium
(NaOH, 2 N) gave carboxylic acid derivatives 21aꢀb. Treatment of 21a with
KOH/NH2OH in THF after activation of the carboxylate afforded the final
hydroxamic acid 22.28, 29 However, we found the carboxylic acid derivative 21a had
higher ER binding affinity than ester 20a and hydroxamate 22 in series I. Although
compounds having carboxylic acid have rarely been studied, Jaouen, et al. prepared a
series of tamoxifenꢀSAHA conjugates in which the carboxylic acid derivative also
had higher ERα affinity compared to hydroxamic acid analogue, and biological assay
results indicated that the carboxylic acid derivative was a more potent inhibitor of
MCFꢀ7 cells.30 Therefore, we wondered whether the replacement of the hydroxamate
unit with the carboxylic acid might also elevate the binding affinity of our OBHS
conjugates. Thus, when the SAHA group was appended onto one of phenol rings on
Cꢀ5 or Cꢀ6 of OBHS (Scheme 3B, Series II), additional SAR investigations were also
focused on the carboxylic acid derivatives 23aꢀr as well as on selected hydroxamate
analogs 24aꢀi that were derived from the carboxylic acid derivatives having better
binding affinities. This hypothesis was verified by the fact that the 23aꢀc, 23f, 23iꢀj,
23m, 23qꢀr had high ER binding affinity in series II; however, when 23aꢀc, 23f, 23iꢀj,
23m, 23qꢀr were transformed into the hydroxamate 24aꢀi with NH2OH after
activation (Scheme 3, Figure 4), ER binding affinity was significantly decreased (see
below).
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 58
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 3. Synthesis of OBHSꢀHDACi conjugates 20aꢀb, 21aꢀb, 22, 23aꢀr and
24aꢀi. Reagents and conditions: (a) neat, 90 °C, 12h; (b) KOH, MeOH, rt, 3h; (c)
ClCO2Et, KOH, THF, 0 °C, 15 min and then NH2OH·HCl, KOH, MeOH, rt, 1h.
10
ACS Paragon Plus Environment

Page 11 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
Table 1. DielsꢀAlder Reaction of Furans 4, 12 and Dienophiles 16aꢀb and 19aꢀr
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 58
1
2
3
4
5
Table 1. Continued
6
Product Yieldb
Conv.a (%)
Entry
Furan
Dienophile
7
8
9
H
N
HO
HO
HO
6
O
O
S
O
O
O
O
O
O
O
O
O
23f
(92%)
S
O
19f
95
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
12
12
O
O
HO
OH
OH
H
N
O
6
O
O
S
O
O
23g
(76%)
S
O
19g
85
HO
H
N
OH
6
O
O
S
O
O
O
S
23h
(83%)
13
O
O
OH
19h
88
O
HO
H
N
HO
HO
O
6
O
O
S
O
O
O
O
O
O
O
23i
(96%)
14
15
19i
19j
99
82
S
O
O
O
HO
H
N
O
F
6
O
O
S
O
O
F
23j
S
O
(78%)
HO
H
N
HO
HO
HO
6
O
O
S
O
O
O
O
23k
(81%)
O
16
17
S
O
O
19k
19l
87
80
OH
HO
H
N
Cl
Br
6
O
O
S
O
O
O
O
O
23l
S
O
O
Cl
(74%)
HO
H
N
HO
HO
6
O
O
O
O
O
O
O
O
O
O
23m
(72%)
S
18
19
S
O
Br
19m
19n
78
88
O
O
HO
H
N
CF3
6
O
O
O
O
23n
(82%)
S
S
O
CF3
HO
12
ACS Paragon Plus Environment

Page 13 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
Table 1. Continued
Conv.a (%)
Product Yieldb
Entry
Furan
Dienophile
7
8
9
H
N
HO
HO
HO
F3C
6
O
O
S
O
O
O
O
O
O
23o
(80%)
O
19o
S
O
20
21
12
87
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
CF3
HO
H
N
O
6
O
O
S
O
O
23p
(93%)
S
O
19p
100
O
HO
H
N
6
O
O
S
O
O
O
S
23q
(90%)
O
19q
19r
22
23
98
O
O
HO
H
N
HO
6
O
O
S
O
O
O
O
23r
(94%)
O
100
S
O
O
HO
H
N
HOHN
O
6
O
S
O
O
24a
(76%)
24
25
S
O
O
19a
93
O
O
O
HO
H
N
HOHN
HOHN
HOHN
O
6
O
S
O
O
O
O
24b
(83%)
S
O
O
19b
19c
96
95
O
O
O
HO
H
N
O
6
O
S
O
O
24c
(79%)
S
O
O
26
27
O
HO
H
N
6
O
S
O
O
O
O
O
O
24d
(76%)
O
S
88
19f
O
HO
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 58
1
2
3
4
Table 1. Continued
Conv.a (%)
5
Product Yieldb
Entry
Furan
Dienophile
6
7
8
9
H
N
HOHN
O
6
O
O
S
O
O
O
O
O
24e
(87%)
19i
19j
28
29
12
90
O
S
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HO
H
N
HOHN
F
6
O
O
S
O
O
O
24f
(73%)
O
S
O
O
F
81
78
92
HO
H
N
HOHN
HOHN
Br
6
O
O
O
O
S
O
O
O
O
24g
(70%)
O
30
31
19m
19q
S
Br
O
HO
H
N
O
O
6
O
O
O
24h
(88%)
S
O
O
S
O
HO
H
N
HOHN
O
O
6
O
O
O
O
O
24i
(85%)
S
19r
32
96
O
S
O
HO
aThe conversion was calculated accounting for the recovered furan 4 or 12. bIsolated
yield by column chromatography purification based on furan 4 or 12.
ER ligands-HDACi conjugates Exhibit Binding Affinity. The binding affinities of
the OBHSꢀHDACi conjugates for both ERα and ERβ were determined by a
competitive fluorometric receptorꢀbinding assay and are summarized in Table 2; the
Ki values of the OBHSꢀHDACi conjugates are also summarized in Table 2.31 These
affinities are presented as relative binding affinity (RBA) values, where estradiol has
an affinity of 100%.
As a global observation, it is noteworthy that the position of suberic acid or SAHA
group in the phenyl ring of OBHS has very significant effects on the binding affinity
of conjugates. In general, the series I of 20aꢀb, 21aꢀb that possessed the suberic acid
group in the phenyl sulfonate moiety demonstrate better binding affinity for ERα than
the series II (the suberic acid group was attached to a phenol ring of OBHS). The
14
ACS Paragon Plus Environment

Page 15 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compound that has the highest binding affinity for ERα and ER subtype selectivity of
all of conjugates is 21a, a compound that possesses a paraꢀsuberic acid group on the
phenyl sulfonate unit. The RBA values of this compound are 12.4 and 0.44 for ERα
and ERβ, respectively, and it has an ERα/ERβ selectivity as high as 28 (Table 2,
entry 3). Compared to the parent compound OBHS (RBA values were 14.4 for ERα,
6.93 for ERβ, α/β was 2.08, measured on a sample newly synthesized for this study),
the conjugate 21a still retained high binding affinity for ERα, and had higher
selectivity (Table 2, entries 3 vs 33). However, compounds 20a and 22 (Table 2,
entries 3 vs 2 and 5), which replaced carboxylic acid of 21a with a methyl ester and
an hydroxamic acid, respectively, both show lower binding affinity for ERα and
reduced subtype selectivity. We are also aware of the substitutional effect that the
position of the suberic acid on the phenyl sulfonate unit has on binding affinity for
ERα. Compound 21b, which possesses a metaꢀsuberic acid group instead of
paraꢀsuberic acid group as in the series I, shows lower binding affinity for ERα than
21a (Table 2, entries 3 vs 4).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As is wellꢀknown, the presence of a phenolic group in ER ligands is crucial to their
binding affinity. OBHS has two phenolic groups, one of which mimics the steroidal
“A ring” present in nature estrogens,32 engaging in strong hydrogen bonds with
residues Glu 353 and Arg 394 and a structured water molecule in ERα. The second
phenolic group in OBHS makes a distinct hydrogen bond with residue Thr 347.
Deletion of the second phenolic OH greatly reduces binding affinity for ERα. Indeed,
all of the series II compounds have decreased binding affinity for ERα; however,
most of the conjugates exhibit moderate binding affinity for ERβ (Table 2, entries
6ꢀ24), and the disposition and the size of substituent on the phenyl sulfonate also
prove to be important factors in determining binding affinity in this series.
In the case of 23a, a compound without a substituent on the phenyl sulfonate, it
shows poor binding affinity for ERβ, with an RBA value not exceeding 1. However,
introduction of a methyl group (compounds 23bꢀc, and 23f, Table 2, entries 7ꢀ8, and
11) resulted in an obvious increase in binding affinity for ERβ, with RBA values
ranging from 3.5 to 5.0, with 23c showing a 10ꢀfold affinity increase for ERβ
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 58
1
2
3
4
5
6
7
8
9
compared to 23a. The position of methyl group, however, has little effect on binding
affinity for both ERs. When the methyl group was replaced with methoxyl group
(analogues 23d, 23e, and 23i), a progressive decrease of RBA values was observed
(Table 2, 23b vs 23d; 23c vs 23e; 23f vs 23i). The paraꢀmethoxyꢀsubstituted
compound (23d) showed lower binding affinities, but the ortho analogue 23i showed
a slight decrease in affinity for both ERs, while the metaꢀmethoxyl analogue 23e
showed a significant drop in binding affinity for ERα, but still had good binding
affinity for ERβ (Table 2, entry 10).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Some interesting things were also observed with hydroxyl analogues 23g, 23h, and
23k. The position of the hydroxy has a remarkable effect on ERβ affinity: The
paraꢀhydroxyl compound 23h is superior to the other two (Table 2, entries 13 vs 12;
entries 13 vs 16) and shows a 16ꢀfold increase over 23a (Table 2, entries 13 vs 6). In
fact, compound 23h has the highest ERβ affinity of all conjugates tested, yet has
modest selectivity between ERα and ERβ.
Comparisons of the ERα and ERβ binding affinity of 23b, 23d, and 23h indicates
that the substituent at the 4ꢀposition of the phenyl sulfonate moiety has a significant
effect on the binding affinity, and halogens substitutions were also evaluated. This
series (23j, 23l, and 23m) showed to an increase in the affinity for ERβ, with RBA
values ranging from 1 to 6, and the fluoro and bromo compounds seem to be superior
to the chloro compound (Table 2, entries 15 vs 17; entries 18 vs 17), whereas the
electronꢀwithdrawing trifluoromethyl substituent (23n) noticeably decreased ligand
binding affinity.
A summary of RBA values of the alternate substituents at 4ꢀposition of the phenyl
sulfonate moiety is presented in Figure 3. These results illustrate that hydroxyl and
halogen substituents have high binding affinity and favor ERβ. Addition of a second
methyl, as in the 2,6ꢀdimethyl conjugate 23p, was found to have about 3000ꢀfold
lower affinity for ERα than the 2ꢀmethyl analogue 23f (Table 2, entries 11 vs 21),
most likely because of unfavorable steric interactions with the pocket. Other changes
to substitutions, such as replacing the phenyl with a naphthyl group, as in compound
23q and 23r, which contained a bulkier substituent than 23p, results in similar
16
ACS Paragon Plus Environment

Page 17 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
binding affinities as those of the methyl analogues for both ERs. Compound 23q and
23r have about 6ꢀ to 12ꢀfold higher affinity than 23a in favor of ERβ. However, when
conjugates 23aꢀc, 23f, 23iꢀj, 23m and 23qꢀr, which had high binding affinities, were
transformed into the corresponding hydroxamates 24aꢀi, all of the hydroxamate
conjugates showed poor binding affinity for both ERs, except 24i, which displays ca
half of RBAs of that of parent compound 23r for ERα and ERβ, respectively (Figure
4). This result also confirmed that the carboxylic acid compound was superior to the
hydroxamate compound in terms of ER binding (Table 2, entries 4 vs 5).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 58
1
2
3
4
5
6
7
Table 2. Relative Binding Affinity (RBA) of OBHS-HDACi Conjugates for ERα
and ERβα
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
18
ACS Paragon Plus Environment

Page 19 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
Table 2. Continued
RBAa (%)
Kib
ER
(
M)
n
6
7
ratio
ER
ratio
ER
ER
Compound
Entry
8
9
H
HO
N
6
O
O
S
O
O
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
91.89
104.02
81.36
23f
0.81
0.88
11
2.98 ± 0.75
3.70 ± 0.21
HO
OH
H
N
HO
HO
6
O
O
S
O
O
O
O
12
101.49
1.25
0.29
23g
3.81 ± 0.04
2.33 ± 0.27
3.35 ± 0.24
8.71 ± 1.02
1.14
0.27
HO
H
N
OH
6
O
O
S
O
O
13
23h
133.05
39.04
HO
H
N
HO
HO
HO
6
O
O
S
O
O
O
O
O
O
O
14
15
23i 1.95 ± 0.30
3.57 ± 0.16
6.83 ± 0.37
0.55
0.49
158.97
92.26
95.24
0.60
0.54
O
HO
H
N
F
6
O
O
S
O
O
23j
3.36 ± 0.49
49.78
HO
H
N
6
O
O
O
O
O
S
23k
0.26 ± 0.05
2.80 ± 0.08
0.09
0.10
1192.30 121.43
16
OH
HO
H
N
HO
Cl
6
O
O
S
O
O
O
O
23l
126.02
125.51
200.01
60.50
1.59
0.48
2.46 ± 0.24
2.47 ± 0.03
0.18 ± 0.03
1.70 ± 0.05
5.62 ± 0.39
1.33 ± 0.16
1.45
0.40
0.14
17
HO
H
N
HO
HO
Br
6
O
O
O
O
O
O
O
O
S
S
18
23m
23n
HO
H
N
CF3
6
O
O
O
O
19
1722.22
255.64
0.15
HO
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 58
1
2
3
4
5
Table 2. Continued
RBAa (%)
Kib
ER
)
n
(
6
7
M
ratio
Entry
ratio
ER
Compound
ER
ER
8
9
H
HO
N
6
O
S
O
O
O
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
O
0.08 ± 0.01
0.11
3875.00 478.87
0.12
20
21
23o
23p
0.71 ± 0.10
CF3
HO
H
N
HO
HO
6
O
O
S
O
O
O
O
O
0.01 ± 0.00
2.55 ± 0.56
0.35 ± 0.14
3.35 ± 0.34
0.029
0.76
971.43
101.49
< 0.19
> 5000
121.57
HO
H
N
6
O
S
O
O
O
O
23q
0.83
22
23
HO
H
N
HO
6
O
O
S
O
O
O
O
23r
1.83 ± 0.16
0.30
0.09
1.00
6.11 ± 0.37
0.11 ± 0.01
169.40 55.65
0.33
HO
H
N
HOHN
6
O
O
O
O
O
O
S
24a 0.01 ± 0.00
24
25
> 5000 3090.90 < 0.62
HO
H
N
HOHN
HOHN
HOHN
6
O
O
O
O
O
O
O
O
O
O
O
O
S
S
S
24b
24c
24d
0.04 ± 0.01
0.07 ± 0.04
0.04 ± 0.00
> 5000
> 5000
0.04 ± 0.00
> 5000
1.00
HO
H
N
6
O
O
O
26
27
< 0.01
> 7
> 1.29
4428.57
> 5000
HO
H
N
6
O
O
O
2125.00 < 0.43
0.16 ± 0.04
0.25
HO
20
ACS Paragon Plus Environment

Page 21 of 58
Journal of Medicinal Chemistry
1
2
3
4
Table 2. Continued
RBAa (%)
5
6
Kib
ER
M )
ER
n
(
Entry
ratio
ratio
Compound
ER
ER
7
8
9
H
HOHN
N
6
O
O
S
O
O
O
O
28
24e
0.04 ± 0.01
0.33
> 5000 2833.33
< 0.57
0.12 ± 0.00
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
O
HO
H
N
HOHN
HOHN
F
6
O
O
S
O
O
O
O
O
O
O
29
30
24f
0.01 ± 0.00
0.03 ± 0.00
0.10
0.21
> 5000 3400.00 < 0.68
0.10 ± 0.01
0.14 ± 0.01
HO
H
N
Br
6
O
O
O
O
O
O
S
S
S
24g
< 0.49
> 5000 2428.57
HO
H
N
HOHN
HOHN
6
O
O
O
O
O
31
32
24h
24i
0.05 ± 0.01
1.01 ± 0.16
0.12
0.30
< 0.16
0.42 ± 0.02
3.32 ± 0.38
> 5000
809.52
HO
H
N
6
O
O
O
O
306.93 102.41
0.33
HO
HO
O
O
O
O
S
33
OBHS
14.40 ± 1.70
2.08
21.53
49.06
2.28
6.93 ± 0.05
HO
aRelative Binding Affinity (RBA) values are determined by competitive flourometric
binding assays and are expressed as IC50estradiol / IC50 compound × 100 ± the range (RBA,
b
estradiol = 100%). Ki values of each conjugate for each receptor were obtained from
the RBA values by the formula Ki = (100/RBA) × Kd. The Kd value of estradiol is 3.1
nM for ERα and 3.4 nM for ERβ, respectively. For details, see Experimental Section.
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 58
1
2
3
4
Figure 3. Graphical exhibits RBA values for alternating substituents at 4ꢀposition of
5
6
the phenyl sulfonate moiety.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Graphical exhibits RBA values for hydroxamic acid conjugates (23aꢀc, 23f,
23iꢀj, 23m, 23qꢀr) and hydroxamate conjugates (24aꢀi).
Transcription Activation Assays. Various OBHSꢀHDACi conjugates were tested
by an ERꢀresponsive luciferase reporter gene assays for their ability to stimulate the
transcriptional activities of ERα and ERβ compared to 17βꢀestradiol (E2). Luciferase
assays were conducted in HEK 293T cells transfected with a widely used 3 ×
EREꢀluciferase reporter. These results are summarized in Table 4, and doseꢀresponse
curves for representative samples and reference drug are shown in Figure 5.
Compounds with agonistic activity were usually classified into three groups,
normal agonist, partial agonist,33 or superꢀagonist,17 based on their maximum efficacy
(Emax) of cell proliferation. The interesting activities are seen in series I. Conjugates
possessing a paraꢀSAHA group at phenyl sulfonate unit, are potent and highly
efficacious ERα and ERβ antagonists; yet, most benzothiophene scaffoldꢀbased hybrid
HDAC inhibitors (termed SERMostats), developed by Thatcher et al, showed little
effect on transcriptional activity,26 and ethynylestradiolꢀHDACi conjugates
(EEDꢀHDACi) developed by Oyelere et al, retained the ERα agonist activity of their
parent compound EED.34 Compound 21a has about a 2ꢀfold improved binding affinity
(Table 2, entries 1 vs 3) for ERα, demonstrated about 216ꢀfold higher potency but
reduced efficacy as an ERα antagonist (Table 3, entries 1 vs 3), compared to
compound 20a. Compared to OBHS, compound 21a has higher efficacy as an ERα
22
ACS Paragon Plus Environment

Page 23 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
antagonist (Table 3, entries 3 vs 33). However, compared to carboxylic acid conjugate
21a and methyl ester conjugate 20a, hydroxamic acid analogue 22 had little effect on
ERαꢀmediated transcription. Interestingly, replacing the paraꢀSAHA with
metaꢀSAHA group gives compounds 20b and 21b, which act as partial ERα and ERβ
agonists.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Compound 23a, which bears the SAHA group on the phenol ring, acts as as ERα
antagonist. In cases of compound 23a, however, modifications of the phenyl sulfonate
moiety (compounds 23bꢀr) result in a wide range of activities on ERα or ERβ.
Compound 23b, which has a paraꢀmethyl on the phenyl sulfonate moiety, acts as an
agonist at ERβ and antagonist at ERα. When the methyl group was changed from the
paraꢀ to metaꢀposition, compound 23c showed improved binding affinity for both
ERs (Table 2, entry 8), but this had little effect on transcriptional activity. The
2ꢀmethyl analogue 23f displayed increased transcriptional activity on ERα compared
to compound 23b. 4ꢀMethoxyl analogue 23d showed lower potency as an ERβ
agonist, along with weaker binding affinity, compared to the 4ꢀmethyl analogue 23b.
These results demonstrate that very small changes to the phenyl sulfonate moiety
can have drastic effects on the potency of these compounds as ERβ agonists. However,
the relationship between RBA values and ERβ transcriptional activity is still unclear.
For example, 4ꢀhydroxyl derivative 23h showed about 16ꢀfold improved binding
affinity for ERβ, and it stimulated ERβ activity with higher potency compared to 23d,
yet the 3ꢀhydroxyl derivative 23g showed about 2ꢀfold weaker affinity for ERβ (Table
2, entries 12 vs 13) and was more efficacious than that of estradiol, being about 2ꢀfold
more potent as an ERβ superagonist than 23h.
Introduction of halogens onto the phenyl sulfonate moiety had obvious effects on
the transcriptional activity of the ER subtypes. The fluoro analogue 23j and the chloro
analogue 23l displayed agonist activity at ERα; actually, 23l profiled as superagonist
on ERα, showing efficacy in ERα 1.5 times greater than that of estradiol (Table 3,
entry 17), while the bromo analogue 23m profiled as antagonist on ERα. The
trifluoromethylꢀsubstituted compounds 23n, and 23o also profiled as ERα antagonist.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 58
1
2
3
4
5
6
Table 3. Effects of OBHS-HDACi Conjugates on the Transcriptional Activities of
Estrogen Receptor α and β
Agonist Modea
Antagonist Modeb
ERα
Eff (%
E2)
ERβ
Eff (%
ERα
Eff (%
E2)c
ERβ
Eff (%
7
8
9
EC50
EC50
IC50
IC50
entry
1
cmpd
20a
20b
21a
21b
22
(ꢁM)
(ꢁM)
E2)
4 ± 4
(ꢁM)
(ꢁM)
E2)
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ꢀ
ꢀ35 ± 1
16 ± 5
ꢀ10 ± 4
51 ± 4
ꢀ15 ± 11
ꢀ13 ± 5
ꢀ6 ± 3
ꢀ28 ± 2
27 ± 7
67 ± 3
ꢀ3 ± 6
77 ± 9
79 ± 5
84 ± 15
38 ± 6
12 ± 1
149 ± 22
ꢀ36 ± 3
ꢀ51± 7
9 ± 1
ꢀ
ꢀ
10.8
10 ± 13
111 ± 13
25 ± 12
122 ± 22
104 ± 29
36 ± 12
36 ± 11
43 ± 1
0.33
76 ± 18
91 ± 17
89 ± 2
2
0.001
3 ± 4
ꢀ
ꢀ
3
ꢀ
ꢀ
ꢀ5 ± 1
0.05
0.16
4
ꢀ
2.1
ꢀ
47 ± 10
3 ± 4
ꢀ
ꢀ
102 ± 5
ꢀ27 ± 3
97 ± 17
111 ± 1
87 ± 15
137 ± 20
100 ± 0
117 ± 14
120 ± 3
23 ± 3
5
ꢀ
ꢀ
1.21
1.08
ꢀ
1.14
6
23a
23b
23c
23d
23e
23f
ꢀ
0.93
0.52
36 ± 10
8 ± 3
ꢀ
7
ꢀ
ꢀ
8
ꢀ
ꢀ11 ± 17
5 ± 1
ꢀ
9
ꢀ
ꢀ
1.1
0.38
8.1
13.4
1.8
ꢀ
115.6
ꢀ
40 ± 27
87 ± 1
ꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
1.16
100 ± 12
24 ± 3
105 ± 1
69 ± 28
65 ± 1
ꢀ13 ± 1
37 ± 0
44 ± 2
ꢀ3 ± 2
ꢀ
ꢀ
0.46
ꢀ
43 ± 1
ꢀ
23g
23h
23i
3.69
108 ± 14
99 ± 12
104 ± 13
99 ± 18
24 ± 4
ꢀ
2.42
ꢀ
11.7
0.016
ꢀ
0.29
78 ± 7
23j
18.0
ꢀ
0.28
85 ± 5
23k
23l
ꢀ
ꢀ
0.14
ꢀ
ꢀ
87 ± 5
0.79
0.58
ꢀ
93 ± 6
ꢀ
188 ± 4
77 ± 15
136 ± 7
102 ± 5
87 ± 12
15 ± 5
23m
23n
23o
23p
23q
23r
24a
24b
24c
24d
24e
24f
ꢀ
1.7
0.79
0.64
75 ± 8
0.25
ꢀ
ꢀ
ꢀ12 ± 3
10 ± 3
ꢀ8 ± 6
10 ± 3
ꢀ
ꢀ
ꢀ
37 ± 15
127 ± 42
14 ± 9
ꢀ
0.32
61 ± 23
5 ± 7
ꢀ
ꢀ
ꢀ
ꢀ13 ± 0.4
19 ± 1
38 ± 4
52 ± 10
29 ± 6
27 ± 11
53 ± 10
49 ± 1
21 ± 9
36 ± 1
ꢀ17 ± 7
0 ± 1
0.13
0.089
ꢀ
ꢀ21 ± 7
20 ± 5
ꢀ28 ± 7
38 ± 2
ꢀ20 ± 5
9 ± 2
0.53
0.002
0.038
0.054
0.203
0.023
0.002
0.173
ꢀ
0.92
45 ± 15
77 ± 5
ꢀ
94 ± 9
ꢀ
ꢀ
ꢀ
92 ± 1
ꢀ
ꢀ
116 ± 3
115 ±6
126 ± 9
164 ± 26
135 ± 1
101 ± 15
88 ± 3
ꢀ
95 ± 2
ꢀ
ꢀ
0.613
48 ± 2
ꢀ
ꢀ
ꢀ
94 ± 6
ꢀ
ꢀ
ꢀ
74 ± 7
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
86 ± 4
24g
24h
24i
1.14
33 ± 9
ꢀ22 ± 8
13 ± 5
60 ± 2
ꢀ
ꢀ
91 ± 6
ꢀ
ꢀ
ꢀ
94 ± 1
ꢀ
0.16
0.014
ꢀ26 ± 5
70 ± 12
7.7
0.581
19 ± 8
OBHS 0.095
ꢀ
ꢀ16 ± 2
aLuciferase activity was measured in HEK 293T cells transfected with 3 ×
EREꢀdriven luciferase reporter and expression vectors encoding ERα or ERβ and
treated in triplicate with increasing doses (up to 10ꢀ5 M) of the compounds. EC50 and
24
ACS Paragon Plus Environment

Page 25 of 58
Journal of Medicinal Chemistry
1
2
standard deviation (mean ± SD), shown as a percentage of 10ꢀ8 M 17βꢀestradiol (E2),
3
4
5
6
b
were determined. IC50 and standard deviation (mean ± SD) were determined in the
percentage of 10ꢀ8 M 17βꢀestradiol (E2) on ERα or ERβ. cERs have considerable basal
activity in HEK 293T cells; compounds with inverse agonist activity are given
negative efficacy values. Omitted EC50 or IC50 values were too high to be determined
accurately.
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 5. Illustrative doseꢀresponse curves for the ERα antagonist effects of 4OHT,
and two OBHSꢀHDACi conjugates 21a and 23q. Efficacy values are the mean ± SD
from three experiments. For details, see the Experimental Section.
Interestingly, comparisons of the ERα antagonist efficacy of 23a, 23b, 23d, 23h
and 23n indicate that substituents at the 4ꢀposition of the phenyl sulfonate moiety
have significant effects. Overall, the electronꢀwithdrawing substituent (CF3) conveys
higher efficacy than the alkyl (methyl) or electronꢀdonating (OMe or OH) groups.
Replacing the phenyl with a bulkier group also resulted in marked effects on the
transcriptional activity. The 2,6ꢀdimethyl substituent (23p) was actually a full agonist
of ERα, while those with bulkier substituents, such as, αꢀnaphthyl and βꢀnaphthyl,
profiled as ERα antagonists. The αꢀnaphthyl substituent had increased ERα antagonist
efficacy with at least 9ꢀfold more potency than 23a (Table 3, entries 22 vs 6), and the
βꢀnaphthyl substituent reduced efficacy, albeit increasing ERα antagonist potency
(Table 3, entries 23 vs 6). However, these bulkier substituents displayed lower
efficacy than the trifluoromethoxy compound (23n). These results are summarized in
Figure 6. The most interesting activities were seen with the hydroxamate conjugates
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 58
1
2
3
4
5
6
7
8
9
24aꢀh. These conjugates had little antagonist potency and efficacy at ERα; however,
most of them profiled as partial agonists at ERβ, and showed greater potency than
their corresponding carboxylic acid derivatives (Table 3, entries 24ꢀ30). Modification
of the 23r by replacing the carboxylic acid group with an hydroxamic acid increased
ERα antagonist potency and efficacy (Table 3, 23r vs 24i).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Antagonist Mode
H
N
HO
R
cmpd
R
%Eff. ER
6
O
O
O
O
O
O
S
23a
23b
23d
23h
23n
H
36
36
40
99
10
Me
OMe
OH
CF3
HO
O
H
HO
N
Ar %Eff. ER
6
O
O
O
O
O
23p
23q
23r
2,6-diMePh
-naphthyl
-naphthyl
127
14
45
S
Ar
HO
Figure 6. Comparisons of ERα antagonistic activities (Eff% of E2) of series II
OBHSꢀHDACi conjugates. The top tables are for changing substituents of phenyl
sulfonate moiety at the 4ꢀposition and the bottom tables for replacing the phenyl with
a bulkier group.
Structural analysis of the origin of enhanced antagonist character of
OBHS-HDACi conjugates. E2 supports transcriptional activation of ERα and ERβ by
stabilizing helix 12 in a position where it forms one side of a hydrophobic groove for
binding transcriptional coactivators. The traditional SERMs or full antagonists have
typically been developed by adding a bulky side group that directly obstructs the
agonist position of helix 12, relocating it out of this position and thereby blocking the
recruitment of transcriptional coactivators.35ꢀ37 By contrast, the oxabicyclic core
derivatives, such as OBHS, have revealed a novel mechanism of antagonism via small
alkyl ester substitutions on the bicyclic core that indirectly modulate the critical
switch helix 12, by interactions with helix 11, a process that we have termed “passive
antagonism”.
Molecular modeling shows that one paraꢀhydroxy phenyl group attached to the
oxabicyclic core of OBHS (Figure 7A) engages in hydrogen bonding with Glu 353,
which mimics the role of Aꢀring phenol of E2. The second phenol points in the E2 11β
26
ACS Paragon Plus Environment

Page 27 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
direction, but is too short to interact directly with helix 12. The key interaction for
antagonism is that of the large, nonꢀpoplar phenyl sulfonate group of OBHS, which
makes strong steric clashes with helix 11 and indirectly modulates the conformation
of the critical helix 12.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Consistent with this model, we find that OBHSꢀHDACi conjugate 21a (Figure 7B)
can similarly form a hydrogen bond between a phenolic hydroxyl group with Glu 353;
moreover, the bulkier sulfonate side chain accentuates this clash with helix 11, thus
giving 21a potent ERα antagonist activity. In contrast, 4ꢀhydroxytamoxifen (Figure
7C) has a bulky side chain that projects between helices 3 and 11, directly displacing
helix 12 from its active conformation and destroying the transcriptional coactivator
binding site. Compound 23q (Figure 7D) mimics the binding orientation of
4ꢀhydroxytamoxifen, with the SAHA group not directly interacting with any helix 12
residues, which is consistent with 4ꢀhydroxytamoxifen. Instead, the SAHA side chain
of 23q forms hydrogen bond contacts with helix 3 (Ser 341, Leu 345, and Leu 346)
which can induce subtle shifts in helix 3 that destabilize helix 12 and destroy the
transcriptional coactivatorꢀbinding site, and again thereby reducing the AF2ꢀmeditated
activity of ERα by an indirect mechanism. Therefore, contact with helix 3 may
represent a novel epitope to generate a full ERα antagonist.
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 58
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 7. Model of OBHSꢀHDACi conjugates bound to ERα and comparisons with
OBHS and 4ꢀhydroxytamoxifen. (A) Computerꢀdeveloped model of oxabicyclic
heptane sulfonate (OBHS) bound ERα (PDB: 3ERD).38 OBHS Hꢀbonds to the
conserved Glu 353 on helix 3. The phenyl sulfonate binds extends between helices 8
and 11. (B) Computerꢀdeveloped model of 21a bound to ERα with the conserved
Hꢀbonding to Glu 353, and the sulfonate moiety extending between helices 8 and 11.
(C) Crystal structure of the ERα LBD in complex with 4ꢀhydroxytamoxifen (4OHT)
(PDB: 3ERT).39 4ꢀOHT forms hydrogen bonds with Glu 353 and Arg 394, the side
chain displaces helix 12. (D) Computerꢀdeveloped model of 23q bound to ERα with
the conserved Hꢀbonding to Arg 394, Ser 341, Leu 345, Leu 346 and Glu 353, and the
sulfonate moiety extending between helices 8 and 11.
Whole cell antiproliferative activity. To evaluate their anticancer activity, all
dualꢀacting conjugates were screened against MCFꢀ7 (hormoneꢀdependent breast
cancer cells), DUꢀ145 (hormoneꢀrefractory, metastatic prostate cancer cells), as well
as one control cell line, healthy kidney epithelial cells (VERO), and the results are
summarized in Table 4. Overall, most OBHSꢀsuberic acid conjugates are much more
28
ACS Paragon Plus Environment

Page 29 of 58
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
potent than lead compound OBHS and corresponding hydroxamic acid conjugates at
reducing cancer cell proliferation, despite the fact that OBHS is equipotent to 4OHT
against MCFꢀ7 cells. This observation indicates that antiproliferative activity of these
OBHSꢀHDACi conjugates against MCFꢀ7 cells derives mainly from inhibition of ER,
and inhibition HDAC can enhance potency against MCFꢀ7 cells. In fact, OBHS does
not possess antiꢀHDAC activity (Table 6, entry 15). Moreover, these carboxylic acid
conjugates show enhanced potency against DUꢀ145, unlike the OBHS, which is
essentially inactive. This also suggests that the antiproliferative activity against
DUꢀ145 of these OBHSꢀHDACi conjugates comes mainly from HDAC inhibition.
Another interesting aspect of antiproliferative activity is that most conjugates
showed significant antiproliferative effects on MCFꢀ7 breast cancer cells, which is
greater than those of DUꢀ145 cells. SAHA was also very effective on MCFꢀ7 cells
and DUꢀ145 cells. This drug, however, showed low cellꢀtype selectivity, in fact, the
majority of HDACi tested clinically exhibit unselective effects on different cellꢀtypes,
and onꢀtarget ineffectiveness.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The compounds 23c, 23g, 23l, and 23q were equipotent to SAHA but had 5ꢀ30ꢀfold
greater selectivity than SAHA against MCFꢀ7 cells (Table 4, entries 9 vs 34; 13 vs 34;
18 vs 34; 23 vs 34). The antiproliferative activity of these conjugates on MCFꢀ7 cells
was also compared with that of the antiꢀbreast cancer drug 4OHT. The compounds
20b, 21a, 23e, and 23r were equipotent to 4OHT (Table 4, entries 3 vs 35; 4 vs 35; 11
vs 35; 24 vs 35), while these four compounds showed about 2ꢀ3 fold greater
selectivity than 4OHT in favor of MCFꢀ7 cells.
Most encouragingly, seven compounds, 20a, 23cꢀd, 23fꢀg, 23l, 23q, exhibited more
potent antiproliferative activity than 4OHT. Specifically, compound 23l shows about
5ꢀfold greater potency than 4OHT (IC50 = 3.3 ꢁM, against MCFꢀ7 cells) and over
30ꢀfold greater selectivity for MCFꢀ7 cells over DUꢀ145 cells (Table 4, entries 18 vs
35), although it showed low binding affinity to both ERs. This suggests that binding
affinity and antiproliferative potency are independent. This conclusion was also
confirmed by compound 23h, which possessed high binding affinity with, however,
no inhibition of either MCFꢀ7 or DUꢀ145 cells.
29
ACS Paragon Plus Environment